Salsola kali
(common saltwort)

Many tumbleweeds (genus Salsola, section Kali) are important roadside, rangeland and agricultural weeds native to much of Europe, North Africa, the Middle East, Central Asia, China and Australia. Species have also been widely introduced accidentally as contaminants of crop and pasture seeds, and tumbleweed has become a widespread weed in Canada, USA, Mexico, Chile, China, Indonesia, Japan, Australia and New Zealand. S. kali and it many confused synonyms (including S. tragus and others which are disputed, such as S. iberica and S. australis) are common agricultural, rangeland and environmental weeds with multiple and important impacts. Major efforts have been made regarding mechanical, chemical and, more recently, biological control.

S. kali is reported as native to much of northern and western Europe, from Norway, Finland, the Baltic states and northern Russia, through Poland, Germany, the low countries, the UK, France, Spain and Portugal (USDA-ARS, 2015).                                                                                

S. kali is reported as introduced and naturalized in Canada, the USA, Mexico, Argentina, Chile, Australia and New Zealand (USDA-ARS, 2015). The Flora of North America Editorial Committee (2015) also records it as invasive in Central and South America and southern Africa.

As S. tragus (here accepted as a synonym of S. kali; see Notes on Taxonomy and Nomenclature), USDA-ARS (2015) records a broader native range than S. kali, encompassing that of S. kali in northern and western Europe, but also extending much further east and south, including the whole Mediterranean region, North Africa (including the Azores and Canaries), West Asia, the Middle East, Central Asia and Siberia, Pakistan, northern India, Nepal and China (USDA-ARS, 2015).

S. tragus is also widely reported as introduced. It is common to abundant in western and central parts of the USA, and found occasionally on eastern and southern coasts, where it is spreading rapidly (Duke, 1983), However, as stated in the Notes on Taxonomy and Nomenclature, uncertainty regarding nomenclature and species limits means that a more detailed review of the history of introduction, in light of a clearer taxonomic a breakdown, would provide valuable insights. In any case, S. kali/S. tragus is highly likely to be more widespread than indicated in the Distribution table.

S. kali appears to be well adapted to a very large range of habitats. Where naturalized, it is found along roadsides, disturbed areas, scrub vegetation, on former sugarcane fields, heath, shrubland and coastal areas (PIER, 2015). It is also a common on fallows, agricultural land and rangeland, especially abandoned fields and overgrazed pastures, shingle and sand behind beaches, and dry gravelly sites in New Zealand. As S. tragus, it is commonly occurs along roadsides and railways, and other dry, stony and sandy areas (Beckie and Francis, 2009).

S. kali is common weed of a very wide range of agricultural crops, as well as grassland plants.

Genetics

The chromosome number is reported to be 2n = 36, but this should be clarified, especially regarding the many synonymous taxa.

Gaskin et al. (2006) called the genus ‘taxonomically challenging’, and undertook morphological, cytological and molecular work to identify an unknown taxon, previously identified as S. tragus, but recently temporarily designated as Type B with unknown origins, present in California and Arizona. They found that it was morphologically very similar, and genetically identical, to S. kali subsp. austroafricana from southern Africa, and distinct from S. tragus. It is unclear if S. kali subsp. austroafricana is native to southern Africa or to the northern hemisphere of the Old World, and further investigations in both regions are needed (Gaskin et al., 2006).

Reproductive biology

The population dynamics of S. australis [S. kali subsp. tragus] were studied in its native range in Western Australia to facilitate the development of improved control strategies (Borger et al., 2009a,b). Two to three cohorts emerged in summer, and in autumn or winter, with an estimated seed production of approximately 100 to 20,000 seeds per plant. Seed viability was low, at 9% for shed seeds and 15% for seeds retained on mature plants. Total annual seed production per unit area was highly variable within sites but was not significantly different between sites (Borger et al., 2009a).

Borger et al. (2009a,b) detailed the taxonomy, morphology, geographical distribution, habitat, growth and development, reproduction, population dynamics, detrimental and beneficial attributes and management (chemical, physical and biological control) of S. australis. See also Crompton and Bassett (1985), discussing S. kali subsp. ruthenica as S. pestifer A. Nels.

Physiology and phenology

As a C4 species, S. tragus (here accepted as S. kali; see Notes on Taxonomy and Nomenclature) is highly competitive in semiarid and arid small-grain cropping systems because of its ability to emerge early, efficiently extract available water from the soil by its extensive root system, and to tolerate heat, drought and salinity. Moreover, the evolution of acetolactate synthase-inhibitor resistance has impacted herbicidal control of the species (Beckie and Francis, 2009).

Environmental requirements

S. kali is native to very wide range of cold, cool and temperate climates. Duke (1983) noted that S. kali ranged from cool temperate desert to steppe to subtropical very dry to thorn forest life zones. Correspondingly, it tolerates a very high range of annual rainfall and annual temperatures (9 to 24°C noted by Duke, 1983). However, subspecific taxa, subspecies, varieties, forms and ecotypes are very likely to be much more specific in their environmental preferences.

In its native range, this highly variable plant is common in coastal beaches and tolerates saline soils, but is also frequently found in inland areas. In the Himalayas, where it is native, it is found at least up to 4,500 m altitude (PIER, 2015).

Much more detailed analysis is required to elucidate specific environmental requirements following taxonomic clarification.

Having been intensively investigated for possible biological control agents, there are many studies on the natural enemies of S. kali in its native range (see Biological control).

In 2005, dying S. tragus [S. kali subsp. tragus] plants were found along the Aegean Sea coast of Greece, with dark brown necrotic lesions on the stems and leaves caused by Colletotrichum gloeosporioides [Glomerella cingulata]. This is thought to be the first report of anthracnose on S. tragus in Greece (Berner et al., 2006). The same pathogen was found on the same species along the Azov Sea in Russia and was the first report of anthracnose caused by C. gloeosporioides on S. tragus in Russia (Kolomiets et al., 2008). In the same region, a further pathogen causing girdling, cracking and dark brown necrotic, canker-like lesions near stem bases, was identified as a Phomopsis, and the conidial state of Diaportheeres, thought to be the first report of stem canker on S. tragus caused by D. eres (Kolomiets et al., 2009).

Animal feed, fodder, forage

• Forage

S. kali is similar to many Salsola species to the point where differences are obscure. Indeed, species limits are not agreed, resulting in confusing taxonomy. This is further complicated by the application of various names, especially where introduced, some of which are likely to have been misapplied.

S. kali is treated here as synonymous with S. tragus sensu stricto (see Notes on Taxonomy and Nomenclature), which is similar to S. paulsenii. Intermediate forms between species are known, and S. paulsenii × S. tragus hybrids have been identified, identical with S. × gobicola. Forms conventionally named as ‘Salsola X’ and ‘S. paulsenii lax form’ with 2n = 54 (e.g. Arnold, 1972; Ryan and Ayres, 2000) are probably also of hybrid origin. S. paulsenii is also noted to be similar in flowers and fruits to S. tragus and S. vermiculata, but it is easily distinguished from them by its shrubby perennial habit and oblong to ovate leaves with round tips (Flora of North America Editorial Committee, 2014). Crompton and Bassett (1985) provided a key that separates the closely related species, S. paulsenii, S. kali, S. collina and S. pestifer. 

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Borger et al. (2009a,b) detailed the chemical, physical and biological control of S. australis [S. kali subsp. tragus]. See also Crompton and Bassett (1985), discussing S. kali subsp. ruthenica as S. pestifer A. Nels.

S. kali is a declared aquatic or terrestrial noxious weed and/or noxious-weed seed in several US states, including Arizona and Ohio as S. kali var. tenuifolia and California and Missouri as S. pestifer (USDA-ARS, 2015). As is common with this species complex, the taxonomy requires clarification.

Cultural control and sanitary measures

A two year study assessed five strategies for controlling S. tragus (here accepted as S. kali; see Notes of Taxonomy and Nomenclature) in Canada (Antill et al., 2012): removal of ungulate grazing; spraying with metsulfuron-methyl; manual pulling; broadcast seeding of native plant species, and integrated weed management combining grazing exclusion, herbicide and seeding. All treatments that included herbicide reduced S. tragus but also removed native forbs. Manual removal reduced weed cover and density but not final biomass, and was considered impractical to implement in large areas. Grazing exclusion resulted in a large decline of S. tragus density and biomass after two years and coincided with recovery of other forbs, suggesting that S. tragus responds to competition and could, therefore, be reduced through management of ungulate grazing pressure (Antill et al., 2012). Matooka et al. (2003) noted in Hawaii that S. kali is grazed, and fences stop the movement of detached plants, and both of which may be reasons why the species has not spread faster than it could have.

Physical/mechanical control

Young plants can be pulled or uprooted, or hoed just below ground level before seed set. Flowers can also be cut before maturity. Mowing tends to cause S. kali to adopt a low habit, but repeated mowing may provde control (PIER, 2015).

In all cases, it is important to establish desirable plants, such as competitive perennial grasses, in disturbed or open areas after any form of control to reduce re-invasion.

Cultivation, however, may promote growth of the weed. A study by Johnson and Fulbright (2008) found S. iberica on fine sandy loams to be 221 times greater on disked plots than on control plots. The authors stressed that land managers should consider soil series when disking for wildlife management, as disking disturbance can clearly worsen invasiveness on certain soils (Johnson and Fulbright, 2008).

Movement control

Attention should be taken regarding the movement of hay from areas where infestation is known. It is uncertain whether seeds survive the passage through livestock guts and therefore whether any quarantine for animals movements is warranted.

Biological control

The gall midge Desertovellum stackelbergi (Diptera: Cecidomyiidae), which attacks S. tragus in Uzbekistan, was identified as a candidate biological control agent for S. tragus in the USA (Sobhian et al., 2003). In a field test conducted in Uzbekistan, Californian S. tragus biotype A, similar to accessions from Uzbekistan, Greece and Ukraine, was the referred host, whereas Californian biotype B was distinct, though also attacked. Since both California biotypes were hosts to the gall midge, further studies on the biology and host specificity of the insect were considered justified (Sobhian et al., 2003).

The obligate biotrophic rust fungus Uromyces salsolae is a candidate biological control agent of S. tragus. Host range tests were conducted in quarantine and only species in the genus Salsola were susceptible to the fungus (Berner et al., 2009).

The eriophyid mite Aceria salsolae was also evaluated as a prospective classical biological control agent of invasive alien tumbleweeds, including Salsola tragus, S. collina, S. paulsenii and S. australis, in North America (Smith et al., 2009). S. tragus was heavily infested by A. salsolae and plant size was negatively correlated to the level of infestation. Although S. kali plants were also infested, plant size did not appear to be affected by the mites. Other nontarget plants were not as suitable for the mite in the field as in previous laboratory experiments, and it was concluded that there would be no significant risk to using A. salsolae as a biological agent in North America (Smith et al., 2009).

Two species of coleophorid moths are also reported to have been introduced to control S. tragus in western USA, where they became widely established; however, they were heavily attacked by predators and parasitoids and have not reduced S. tragus populations.

Cosmobaris scolopacea is a weevil distributed in Eurasia and North America, generally associated with the Chenopodiaceae family of flowering plants. The larvae feed and pupate within stems. Preliminary host range testing in Italy confirmed the presence of a highly divergent Sicilian lineage of C. scolopacea that was reared only on S. kali, indicating that this type requires further testing as a biological control agent (Cristofaro et al., 2011).

Colletotrichum gloeosporioides f. sp. salsolae is a facultative parasitic fungus being evaluated as a classical biological control agent of S. tragus in initial host range determination tests (Berner et al., 2012). Greenhouse tests showed that it is specific to Salsola spp., and that pathogen and inoculation procedures offer a low-cost solution to S. tragus infestations (Berner et al., 2014).

Despite the large number of identified potential biological control agents and host specificity testing, there appear to be few records of actual releases or more widespread field testing of the above.

Chemical control

Glyphosate can control S. kali if applied before seed set, whereas 2,4-D may actually cause S. kali to become tough and leathery, producing a plant that is more difficult to manage. It has also been observed that some S. tragus are showing resistance to sulfonylurea and trazine herbicides in western USA (PIER, 2015), though it is noted as very sensitive to hormone-type dicot herbicides (Motooka et al., 2003).

In northwestern USA, where it is a severe postharvest problem, S. tragus [Salsola kali subsp. tragus] is controlled either by tillage or applications of various herbicides. Paraquat plus diuron gave effective (>90%) control, whereas glyphosate plus 2,4-D provided unacceptable (≤75%) weed control. The use of laser controlled application led to a 42% reduction in chemical use compared with broadcast applications with a corresponding cost (Young et al., 2008).

In cotton fields, application of 2,4-D effectively controlled S. tragus, and could be applied close to the plant without injuring cotton, whereas dicamba and diflufenzopyr plus dicamba were less effective as size of the plant increased and both injured cotton regardless of interval between application and planting (Everitt and Keeling, 2007). Many studies have looked at various herbicide treatments and the economics of timing, application rate, and impacts on crop yields, including in sunflower (Godar, 2013) and beans (Arnold et al., 2012).

IPM

Borger et al. (2009b) concluded that effective weed management depended on: reducing seed dispersal; burning all senesced, mobile plants in late autumn, and herbicide control of the largest cohorts of S. australis in summer and autumn. Together, this reduced population growth rate. Using a model, they predicted that this control strategy resulted in a 66% chance of the population becoming extinct over 25 years.

Salsola section kali requires a detailed literature review and thorough investigation, beginning with a full clarification of taxonomic limits of defined taxa in the native range and, where introduced, corresponding populations and the elucidation of parents in the case of hybrid populations. Only when this has been achieved and agreed can further advances be made on control. It is evident that different taxa have quite different environmental requirements and respond quite differently to different treatments. A full review of control methods, especially chemical and biological control, appears warranted.

30/04/15 Original text by:

Nick Pasiecznik, consultant, France